Aribot by Ayurak

What teams use Aribot for

41 use cases, each with a one-page summary and a carousel you can share. 6 case studies: two from a live customer, the rest from running Aribot on our own platform, labelled as such.

Case studies

Case study · Ayurak Trust

Eleven plant managers asked. Eleven assessments back.

The group's security governance team runs its cybersecurity self-assessments for plants and suppliers on Ayurak Trust. Every plant manager it asked completed the operational-technology assessment, and every supplier it asked about process-control systems sent theirs back complete.

11 of 11
plant managers completed the OT self-assessment (2023 round)
6 of 6
supplier OT and process-control assessments returned complete

Case study · Aribot threat modeling

From a request form to a signed-off threat model

The security governance team wanted every new system to get a threat model without the team becoming the bottleneck. Aribot now takes the request, works out which frameworks apply from the answers, opens the work in ServiceNow and tracks it to sign-off.

8
frameworks the intake can scope automatically, including ISO 27001, ISO/IEC 42001, SOC 2 and NIST CSF 2.0
15 min
between ServiceNow and Aribot ticket reconciliations, on top of instant webhooks

Case study · Shadow AI · Customer zero

Our own code called five AI providers and 25 models

We pointed Aribot's Shadow AI discovery at our own codebase. It found 134 places where the code calls an AI model, spread across five providers and 25 models. Set next to our AI cost report, the inventory exposed a gap: calls to two of those providers were being recorded as free.

134
distinct places in our code that call an AI model
5
AI providers in use, across 25 different models

Case study · AI governance · Customer zero

Seventeen agent requests refused, each with a reason on record

Every AI agent that calls Aribot on our own platform goes through the same gateway our customers use. The gateway decides each request, writes the decision to a hash-chained ledger and refuses what the agent should not do. Of 40 governed requests, it refused 17, and each refusal names its reason.

23
agent identities registered with the gateway
17 of 40
agent requests refused, each with a recorded reason

Case study · Self-healing · Customer zero

Two out of three fixes never touched production

Aribot's self-healing proposes a fix for a cloud or design finding, runs it as a dry run, waits for approval and checks the result after it applies. On our own platform, 49 of 73 remediation runs stopped at the dry run. Of the 24 live runs, eight were applied and verified and three failed. Thirteen from February never finished; we checked every resource they targeted before closing them.

49 of 73
remediation runs that were dry runs only, February to September 2026
8
live fixes applied and then verified against the resource

Case study · AI cost · Customer zero

150 paid AI calls a night, down to a handful

Our own AI cost report showed the same spend every night at 04:30. Aribot traced it to one scheduled job, showed which calls were wasted, and after two small fixes the first two nightly runs made 4 and 7 model calls instead of about 150.

~150 → 4–7
paid model calls in the nightly cost run, before the fix and on the first two nights after
$4.85 → $0.26
cost of that run on the last night before, and the higher of the first two nights after

Use cases

Platform & Strategy · 9

From architecture diagram to a fix you can check

Aribot follows one path from your architecture to an approved, re-checked fix, and keeps the evidence for each step at the step that produced it.

Turn a plain-language brief into a threat-modelled design

Describe the system you want in plain language, and Aribot turns it into a cloud architecture and a threat model that stay linked to the live estate.

Show the board exposure in currency, with coverage

Aribot prices security exposure per system and shows, beside every score, how many controls were tested, so the board can see what the figure rests on.

See what each AI agent can reach and did

Aribot finds every AI agent that holds a credential and shows the access it was granted beside the access it used, with a verdict on every action.

Fix cloud findings in bulk behind an approval gate

Aribot turns routine cloud fixes into dry runs behind policy gates, and a finding closes only when the cloud confirms the change.

Close a finding only when the system confirms it

A finding stays open until the affected system has been re-checked and agrees, so your remediation record lists confirmed changes.

Cut machine access down to what is used

Aribot shows what each agent or machine identity holds beside what it has used, so access reviews rest on evidence and unused access can be removed.

Report untested controls as untested

Aribot keeps unmeasured as its own visible state, so every score carries its coverage and an untested control never counts as a pass.

What Aribot is building next, by buyer problem

A preview of the problems Aribot is working on next, such as proving a fix worked and matching machine access to what is actually used.

Cloud Security Platform · 14

Stop AI spend at the limit, automatically

Aribot checks AI spend against your billing export hourly and switches off paid AI capacity at the limit, with cloud cost in the same view.

Find recoverable cloud spend and separate modelled from banked

Aribot models the cloud spend you could recover and keeps that figure apart from the savings you have actually made.

One view of findings across AWS, Azure and GCP

A single operations view of alerts, resources and posture across your AWS, Azure and GCP accounts, so triage starts in one place.

Connect cloud accounts read-only and see which failed

Connect AWS, Azure and GCP accounts with read-only access and see which ones Aribot could actually read, so an unread account never passes as clean.

A posture score that shows which accounts it covers

Scan results across connected accounts, where the headline score states how many accounts it was computed over and unscanned accounts stay on the list.

Triage multi-cloud findings without losing the unrated ones

Findings from every connected cloud in one list, with severities from the source and a separate band for findings that arrived with none.

Framework scores that show how many controls were assessed

Aribot maps findings to the controls in each framework you report on, and every framework score states how many controls were actually assessed.

AI-drafted fixes that wait for your approval

Aribot drafts a fix for each finding with the model that wrote it on record, and nothing changes until a person approves the dry run.

Ask about your estate in plain English

Ayurak Intelligence answers questions about your own estate with its sources cited, and any change it proposes waits for a person to approve it.

Threat intelligence mapped to the systems you run

Aribot joins MITRE ATT&CK techniques and CVE data to the components and resources you have connected, so a new threat shows which systems it reaches.

Step-by-step remediation plans with outcomes recorded per run

Aribot generates step-by-step remediation plans for findings across AWS, Azure and GCP, and each plan's track record comes from its own executions.

An execution log that includes work that never started

Every remediation, scan and policy run with its outcome, including jobs that were scheduled and never began, exported as one trail.

Kubernetes findings in the same view as your cloud

Cluster findings such as privileged containers or automounted credentials land in the same violations view as the rest of your estate, scoped to cluster resources.

An audit trail the auditor can verify

Every live remediation is written to an append-only ledger, hash-chained and signed, and anyone with access can verify the chain on demand.

Threat Modeling · 8

Architecture, threats, controls and cloud evidence on one canvas

One canvas joins your architecture, the threats on each component, the controls that answer them and what your live cloud says about those controls.

Code findings tied to the line that caused them

Paste code and get findings mapped to STRIDE and MITRE ATT&CK, each tied to the line that produced it and paired with a recommended fix.

Threats per component, straight from your architecture

Aribot reads your architecture and attaches threats to each component, mapped to STRIDE and MITRE ATT&CK, so design review keeps up with delivery.

A secure architecture diagram in one click

Describe the system, pick the cloud and the standards you answer to, and Aribot draws the architecture with the security controls in place, then checks it.

Replay an attack path across your architecture

Aribot plays an attack chain step by step across your architecture, labelling each move with its MITRE ATT&CK technique and the component it reaches.

Test a design change against a mirror first

Model a proposed change against a mirror of your connected estate and see cost, dependencies, downtime risk and control status before anything moves.

Threat model as a team, with disagreements on record

Several people edit one architecture at once, Aribot derives threats on each change, and disputed threats stay on the model with the reason attached.

Let AI assistants call Aribot through one governed gateway

Point any MCP client at the Aribot server, and its calls pass through the same gateway and permission checks as every other agent action, with write calls metered.

By industry · 10

DORA evidence that stays current between audits

Aribot builds the ICT risk picture from your architecture and keeps the evidence current as systems change, so the third-party register is never a quarter-end project.

Every payment API in your code, ranked by risk

Aribot lists every endpoint in your code, ranks the risky ones and maps the fixes to PCI DSS, so security review keeps pace with releases.

Pass the enterprise security questionnaire with evidence

Aribot scans your repositories and pipelines and turns the fixes into the SOC 2 evidence an enterprise buyer's questionnaire asks for.

Answer the question every buyer asks about your agent

Aribot shows what your agent can touch, keeps a verifiable record of every action, and threat-models the agent itself.

Find the AI already running, then govern it

Aribot finds the AI already running in your code and cloud, then puts every agent under least-privilege access with a record your auditors can check.

A small team fixing the right cloud gaps first

Aribot scans your cloud read-only and maps every gap to HIPAA and GDPR, so a lean security team works on what matters most.

A threat model and an SBOM for every release

Aribot produces the design-stage threat model and the SBOM regulators expect, from your architecture and code, and keeps both current release to release.

Plant and cloud attack paths, mapped to NIS2

Aribot models your plant and cloud architecture, shows the attack paths that matter and maps the controls to NIS2, so management can show it is in control.

A documented design review in a day, not a tender cycle

Aribot turns a new service's architecture into a documented threat model and control map, so the security review is ready before go-live.

Sell the judgement, not the legwork

Aribot does the first pass on threat models and cloud reviews in minutes, so your consultants spend their days on the advice clients pay for.